A SVM regression predicting model for indentation depth of welding spot based on digital image processing

Jin Zhang; Peng-xian Zhang

doi:10.1117/12.956422

8 June 2012 A SVM regression predicting model for indentation depth of welding spot based on digital image processing

Jin Zhang, Peng-xian Zhang

Proceedings Volume 8334, Fourth International Conference on Digital Image Processing (ICDIP 2012); 83342H (2012) https://doi.org/10.1117/12.956422
Event: Fourth International Conference on Digital Image Processing (ICDIP 2012), 2012, Kuala Lumpur, Malaysia

Abstract

The indentation depth of resistance spot welding joint is closely related to its quality, and the digital image of its surface is used as an information source. An evaluating algorithm of artificial intelligence for the indentation depth is put forward. Firstly, through analyzing characteristics of images on the surface of spot welding joints, the first ring area S₁, the second ring area S₂, total area S, the area ratio coefficient between total area and first ring area K₁, and the area ratio coefficient between total area and second ring area K₂ are extracted as evaluation factors of indentation depth. At the same time, S₂, S, K₁ are selected as characteristic parameters of the indentation depth based on the correlation analysis between the evaluation factors and the indentation depth. Secondly, a support vector machine (SVM) predicting model of the indentation depth is established. The model selects the parameters S₂, S, K₁, welding current I, and electrode pressure F as the input vector and selects the actual indentation depth h_T of welding spot as the target vector. Test results are shown, the correlation coefficient are 0.9958 between model prediction values and actual measured values. The indentation depth of welding spot can be predicted by means of the SVM evaluating algorithm.

Citation Download Citation

Jin Zhang and Peng-xian Zhang "A SVM regression predicting model for indentation depth of welding spot based on digital image processing", Proc. SPIE 8334, Fourth International Conference on Digital Image Processing (ICDIP 2012), 83342H (8 June 2012); https://doi.org/10.1117/12.956422

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available